Browsing by Author "Palanisamy, P"

Now showing 1 - 9 of 9
  • No Thumbnail Available
    Item
    Automated Binary and Multiclass Classification of Diabetic Retinopathy Using Haralick and Multiresolution Features
    (Institute of Electrical and Electronics Engineers Inc., 2020-03-10) Gayathri, S; Krishna, Adithya K.; Gopi, Varun P.; Palanisamy, P
    Diabetic Retinopathy (DR) is considered as the complication of Diabetes Mellitus that damages the blood vessels in the retina. This is characterized as a serious vision-threatening problem in most of the diabetic subjects. Effective automatic classification of diabetic retinopathy is a challenging task in the medical field. The feature extraction plays an eminent role in the effective classification of disease. The proposed work focuses on the extraction of Haralick and Anisotropic Dual-Tree Complex Wavelet Transform (ADTCWT) features that can perform reliable DR classification from retinal fundus images. The Haralick features are based on second-order statistics and ADTCWT reliably extracts the directional features in images. The proposed work concentrates on both binary classification as well as multiclass classification of DR. The system is evaluated across various classifiers such as Support Vector Machine (SVM), Random Forest, Random Tree, J48 classifiers by giving input image features extracted from the MESSIDOR, KAGGLE and DIARETDB0 databases. The performances of the classifiers are analyzed by comparing specificity, precision, recall, False Positive Rate (FPR) and accuracy values for each classifier. The evaluation results show that by applying the proposed feature extraction method, Random Forest outperforms all the other classifiers with an average accuracy of 99.7% and 99.82% for binary and multiclass classification respectively.
  • No Thumbnail Available
    Item
    Automated weld defect detection using gated attention and squeeze and excitation fusion U-Net
    (Nature Research, 2025-11-18) Popuri, Jaya Ratnam; Nitish, N. Hari; Kumar, Dheeraj Dhruva; P. Gopi, Varun; Palanisamy, P
    Welding is a fundamental process in manufacturing and construction, playing a pivotal role in joining materials by melting and fusing them. This versatile technique finds widespread application across diverse industries, creating robust and structurally sound components. Regular inspection and continual oversight of the welding process are imperative. Manual inspection methods, however, prove inadequate in addressing the ever-growing demand for safety and precision. In recent years, integrating deep learning and computer vision technologies has revolutionized the welding field by enhancing and automating inspection and safety protocols. This paper proposes DeepFuse WeldNet, an automated Weld detection Framework that employs a novel Gated Attention Squeeze and Excitation Fusion U-net (GASEUNet) for weld defect segmentation. Our research focuses on enhancing the accuracy of the model for more precise identification and measurement of weld porosity within images. The model provides a significant boost in segmentation accuracy that is better than existing methods. Our experimental results demonstrate the effectiveness of GASEUNet, with an impressive Jaccard Coefficient (JC) of 98.12. The proposed DeepFuse WeldNet also consists of a fully automated Weld Inspection Framework that classifies the weld defects as safe or unsafe by comparing the measurements to the standards provided by the American Welding Society (AWS)
  • No Thumbnail Available
    Item
    Classification and Indirect Weighing of Sweet Lime Fruit through Machine Learning and Meta-heuristic Approach
    (Bellwether Publishing, Ltd., 2021-05-03) Phate, Vikas R.; Malmathanraj, R; Palanisamy, P
    In the past few decades, both academicians and industries have shown interest toward the agricultural post-harvest operation aiming to reduce the post-harvest losses. In order to assist farmers in post-harvest decision-making some effective and innovative methodological frameworks are required. The fruit weight measurement is of prime importance in many food processing industries during sorting, grading, and packaging. In this work, different Support vector machine (SVM) classifiers as well as weighing models developed using the optimized adaptive neuro-fuzzy inference system (ANFIS) coupled with a computer vision system are proposed. More precisely, the weighing models based on the hybrid ANFIS approach using two well-known optimization algorithms are analyzed. In the first approach, a series of GA-ANFIS models have been evaluated for different population size. In the later approach, different PSO-ANFIS models have been evaluated by varying the most influential parameters. The comprehensive self-built color image database has been used for both calibration and validation of the models. From an economic point of view, this indirect way of weighing fruits may be useful to fruit growers and traders in deciding the market depending on the fruit size and weight before packaging. The result shows the higher reliability and prediction capability of the proposed meta-heuristics (GA-ANFIS) model in estimating the weight of sweet lime fruit.
  • No Thumbnail Available
    Item
    CNN driven nutrient deficiency detection in plants using real-world leaf images
    (Learning Gate, 2024-09-11) Sathyan, Anish; Palanisamy, P
    In agriculture, our farmers currently lack efficient methods to classify and identify plant nutrient deficiencies or diseases, leading to ineffective remedial measures. However, the potential for early detection of plant nutrient deficiencies using colour gradients or leaf patterns, coupled with the advancements in Image Processing, Deep Learning (DL), and Artificial Intelligence (AI), offers a promising future. This paper presents deep learning-based methods to segregate nutrient deficiencies using leaf images, a potential game-changer for the industry. Our research has the potential to revolutionize plant nutrition management, inspiring a new wave of efficient and effective practices in agriculture. We analyzed the International Plant Nutrition Institute (IPNI) dataset and applied an image augmentation mechanism to enhance the training process with the available dataset. By utilizing Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and learning algorithms, we minimised loss and improved accuracy. The model, incorporating over 1.5 million parameters, has shown promising results in predicting deficiencies accurately. We propose an SGD-based training mechanism to balance localization and classification tasks, a step towards a more efficient and effective approach to plant nutrition management.
  • No Thumbnail Available
    Item
    Confidence-Aware Cascaded Learning for Real-Time Weakly Supervised Anomaly Detection in Surveillance Video
    (Institute of Electrical and Electronics Engineers Inc., 2026-04-01) Mohanarangan, K; Palanisamy, P
    Weakly supervised video anomaly detection (WS-VAD) has advanced significantly, yet most approaches remain tailored to long, untrimmed streams and fail to meet the latency and reliability demands of real-world surveillance. In asset-protection scenarios, where motion-triggered clips must be processed in real-time, existing methods often compromise either accuracy or efficiency. We present a Confidence-Aware Sequential Routing Ensemble (CASRE), a cascaded framework designed for Clip-Level Weak Supervision (CLWS). CASRE integrates a temperature-scaled ResNet3D (TDR3D) primary detector with a novel Temporal-Scale Delta Loss to enforce confidence separation. Only low-confidence clips are routed to a transformer-based MViTv2 fallback, and final predictions are obtained through calibrated probabilistic fusion. This selective routing ensures that high-capacity models are invoked for fewer than 6% of clips, while the majority benefit from lightweight inference. Extensive evaluation on UCF–Crime, UCSD–Ped1/Ped2, and the newly introduced SSAD dataset, a real-world benchmark curated for asset-protection surveillance, demonstrates that CASRE achieves up to 98.64% accuracy and 99.91% AUC, while maintaining sub-0.03 s inference for most samples. By combining confidence calibration, selective routing, and heterogeneous backbone fusion, CASRE establishes a deployable WS-VAD paradigm that balances accuracy, robustness, and computational efficiency, making it well-suited for large-scale surveillance deployments under strict operational constraints.
  • No Thumbnail Available
    Item
    Detection of nutrient deficiency in plants using radial basis function neural network with a modified adaptive Kalman filter
    (Horizon e-Publishing Group, 2025-10-23) Sathyan, Anish; Palanisamy, P
    The agriculture sector plays a significant role in the national GDP. Global agricultural industries employ advanced techniques for the early detection and characterisation of plant nutrient deficiencies to enhance crop yield. In most agriculture-dependent countries, economic growth is determined by the gross yield of entire plantations. Plants exhibit noticeable symptoms such as changes in leaf colouration, spotting, or altered growth patterns. These symptoms, including yellowing (chlorosis), browning, or stunted growth, indicate potential nutrient deficiencies. Monitoring these variations enables early detection of deficiencies before they significantly affect plant health. Early detection of nutrient deficiencies plays a pivotal role in achieving a speedy recovery from the disease, thereby assuring sufficient improvement in the plant’s yield. This article proposes a method for the early detection of nutrient deficiency in plants based on variations in leaf colour and pattern. It uses 54305 data points with three versions of the same images, differing in colour, greyscale and segmentation. Initially, image preprocessing is carried out using Fuzzy Histogram Equalization (FHE) and a modified adaptive Kalman filter to achieve better results. The preprocessed image datasets are then used to train the proposed Radial Basis Function Neural Network (RBFNN). The proposed RBFNN model demonstrates significant improvements (1.18-16.47 %) in accuracy, precision, specificity, sensitivity and F-score compared to conventional models.
  • No Thumbnail Available
    Item
    Disagreement-aware late fusion of 3DCNN and MViT for clip-level video anomaly detection
    (Springer Science and Business Media B.V., 2026-06-30) Mohanarangan, K; Palanisamy, P
    We propose Adaptive Late Router with Temperature Scaling (ALeRT), an inference-only, disagreement-aware late fusion framework for clip-level video anomaly detection. ALeRT combines two complementary experts: MMAnet, a lightweight multi-scale 3D CNN for fine-grained motion dynamics, and MViT, a Kinetics-pretrained multiscale Vision Transformer for long-range spatiotemporal modeling, without requiring joint retraining. Model outputs are calibrated using temperature scaling and fused via a Product-of-Experts (PoE) to reinforce agreement, while disagreement is resolved using confidence margins and per-class Positive Predictive Value (PPV). Temporal consistency is further enhanced through exponential moving average (EMA) smoothing. The framework is evaluated across multiple random seeds, reporting mean performance with low variance, demonstrating strong robustness and generalization. Experiments on LPS (Loss Preventive Surveillance), UCF-Crime-Clips (UCF-CC), and Virat2-RC (ReClassified) datasets show consistent improvements over individual backbones. Notably, ALeRT achieves these gains with only 3.1 ms fusion overhead per clip, enabling real-time inference (5.7–6.0 events/s, maximum of 300 FPS(FPS (frames per second) indicates the number of video frames processed per second.)) on commodity GPUs. These results highlight the effectiveness of disagreement-aware fusion for scalable and deployment-ready video anomaly detection.
  • No Thumbnail Available
    Item
    Prediction of Gate In Time of Scheduled Flights and Schedule Conformance using Machine Learning-based Algorithms
    (Embry-Riddle Aeronautical University, 2020-11-18) Sahadevan, Deepudev; Palanisamy, P; Gopi, Varun P.; Nelli, Manjunath K.; Asok kumar, A.K.
    Over the past several years, the air transportation system has encountered frequent rises in air traffic demand, especially with the introduction of budget airlines. In recent Indian-aviation submits, IATA has presented that by 2026 India is expected to be the third-largest air transport market in the world (International Air Transport Association, 2018) from its current 7th place. Before pandemic COVID-19, increases in air traffic in trend were predicted such that the current air traffic system will not be able to cater the projected air traffic demand of the near future due to the capacity constraints of airport and airspace. In order to overcome this backdrop, there arises a need for the of modernising the air transportation system. Prompt initiatives are being taken by several countries, including India, to develop the future air transport system that will be more robust, predictable, and reliable than today’s one. Introduction of Central Air Traffic flow management(C-ATFM) in India was one of the steps taken towards achieving this objective.
  • No Thumbnail Available
    Item
    TriPath3DNet: an efficient real-time model for multi-class classification in real-life surveillance videos of fixed duration
    (Nature Research, 2026-06-17) Mohanarangan, K; Palanisamy, P
    This paper presents TriPath3DNet, a novel, efficient, and interpretable 3D CNN architecture designed for real-time, multi-class classification of short, motion-triggered surveillance video clips under challenging real-world conditions–including occlusion, variable lighting, adverse weather, and subtle low-motion anomalies such as loitering or zone intrusion. TriPath3DNet integrates three complementary temporal pathways–short-term motion, long-term context, and Temporal Difference Encoding (TDE)–into a lightweight ResNet3D (R3D)-18 backbone to jointly model transient dynamics and sustained activities. Evaluated on four datasets–including the newly curated Virat1-RC, Virat2-RC, UCF-Crime, and our proprietary In-House Dataset (IHD)–TriPath3DNet achieves state-of-the-art or near state-of-the-art performance, with up to 95.37% accuracy, 99.42% AUC, and an inference latency of 129 to 137 ms per 50-frame clip (approx 2.6 ms per frame) on an 11 GB GPU, using only 33.46 M parameters. Notably, it outperforms both CNN- and transformer-based baselines–including MViTv1, MViTv2, and VideoSwin–by significant margins, especially on anomaly-dense benchmarks like UCF-Crime, where most vision transformers struggle. While MViTv2 achieves slightly higher accuracy on IHD (91.87% vs. 89.46%), TriPath3DNet delivers substantially better AUC (98.07% vs. 90.96%), indicating superior calibration for critical anomaly detection. Ablation studies confirm that each temporal branch contributes meaningfully to performance, and Grad-CAM visualizations demonstrate spatially precise and temporally coherent attention maps. By aligning architectural design with edge–cloud deployment constraints and the operational realities of industrial surveillance, our work bridges the gap between academic research and real-world video analytics.